The mobile internet was a simpler infrastructure to design than the one that will be needed for the smart city, smart grid, smart health and smart transportation. Smartphones are homogeneous with relatively powerful processors and batteries driving transmission and reception. Designed to bring the internet to smartphones, 3G and 4G networks could be simpler. But IoT devices will span a range of heterogeneous designs.
The range of heterogeneity of the IoT is defined today by autonomous vehicles, which have thousands of sensors powered by high-capacity batteries that frequently communicate at high speed and low latency to simple sensors. Those sensors are powered by ambient power sources, sending a few infrequent bytes to communicate state (on/off, temperature, vibration amplitude and phase, etc.).
Add to the homogeneity/heterogeneity comparison, wireless design only had to meet the largely common communications needs of consumers and private and public enterprises. The next smart infrastructure, however, will simultaneously interconnect diverse IoT devices and smartphones. It will need to be extensible to more than 10 times the number of device and 1,000 times the bandwidth of today’s wireless networks. In addition, radio coverage will need to be denser with greater signal penetration through obstacles.
A team of seven researchers from top U.S. technical universities published a research paper funded by the National Science Foundation that defines the impact of IoT heterogeneity on the future of smart infrastructure. The purpose of the paper is to recommend increased research and development in the design of smart wireless infrastructure. While the researchers are not critical of the wireless industry overall, they said:
“Designing a new wireless infrastructure that truly meets the needs of communities is extremely challenging and demands a vision beyond what the wireless industry alone can provide.”
4 smart infrastructure areas that need more R&D
Specifically, the researchers recommend increased R&D in the following four areas:
1. Flexible infrastructure
Software-defined networking, including software-defined radios for flexible infrastructure, needs to be considered.
More research is needed for machine learning and big data analysis for dynamically analyzing wireless signals and performance monitoring to increase efficient use of infrastructure and spectrum.
Optimum distribution of computation, storage and communication resources, ranging from the highly cost- and energy-efficient cloud to the hyper-responsive fog computing mobile edge needs to be considered.
Virtualization of the wireless networks, decoupling network infrastructure ownership from services for greater flexibility and cost-effectiveness and to isolate critical traffic for smart infrastructure, must be considered.
A limited public access option to this infrastructure for businesses that are expected to want to offer the service like free Wi-Fi today should be part of the architecture.
2. High-bandwidth and agile wireless technologies
High-speed routers with extreme bandwidth that cluster communications resources to meet dynamic low latency application demand will be needed. Design should reduce traffic on fronthaul networks. Fronthaul networks are fiber networks connecting the edge to the central radio access network (RAN) backhaul network.
The researchers recommend further study of extremely high-bandwidth radio access technologies to support software-defined networks for practical mobile environments. Areas of focus cited are ultra-dense cells, millimeter wave, teraHertz band, massive MIMO (multiple-input and multiple-output) and carrier aggregation.
The intermittent connectivity to work with low-power devices that may be temporarily unreachable to save energy should be part of the architecture.
3. Reliability, security and resilience
Security solutions that authenticate and are hardened against attack and privacy leaks should be included in the architecture for non-stationary mobile and for IoT devices that use multiple access technologies and are capable of optimized simultaneous communications. There’s a need for new programming tools and software stacks to lower the barrier for creating effective wireless services that have good performance, reliability, energy efficiency, etc.
Further study of energy-efficient wireless sensors and actuators that report data and effect change in the physical world on the IoT and smart infrastructure will need further study. New techniques for energy harvesting to power IoT devices without batteries are included in this study.
Seamless integration of heterogeneous access technologies (Wi-Fi, cellular, wired, etc.), including using multiple technologies at the same time for better coverage, performance and reliability.
Disaster scenarios that provide low-bandwidth reliable communication should be designed into the architecture for critical IoT and smart infrastructure.
The researchers’ assert:
“It is easy for technologists, engineers, and researchers to dream up high-tech ‘solutions,’ but [it is] much harder for these professionals to work with local authorities and organizations, such as law enforcement, public housing, transportation, schools, government and philanthropic organizations, to find solutions that will raise the human condition.”
Towards this end, they recommend partnerships with many organizations.
Use of dynamic spectrum and unlicensed spectrum should be tested in partnership with the FCC and NIST.
Philanthropic organizations such as the 100 Resilient Cities project created to deal with the shocks and stressors that weaken the fabric of cities can contribute a sustainable urban perspective to research.
Open-source software and open interfaces/standards for programming and managing wireless networks, created in partnership with existing and new standards organizations, would enable innovation and interoperability.
Academic partnerships to train students, government employees and workers from other fields in the use of wireless technologies would provide the development resources to create societal and business IoT solutions.
Partnerships with academia are needed to help train students and government employees (e.g., law enforcement) and retrain workers in many fields, as well as to create and use wireless technologies to build solutions to address important societal and business problems.